Develop deep-learning–enhanced hadronic models to improve nuclear reaction cross-section predictions for particle therapy and space radiation protection, integrating data-driven models into Monte Carlo frameworks and validating with ion-beam experiments.